Componential Analysis of Kinship Terminology

A Computational Perspective

V. Pericliev

• Publisher: Palgrave Macmillan
• Year: 2013
• ISBN:
  9781137031181
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

This book presents the first computer program automating the task of componential analysis of kinship vocabularies. The book examines the program in relation to two basic problems: the commonly occurring inconsistency of componential models; and the huge number of alternative componential models.

1

Introduction: the historical background

This chapter looks at componential analysis of kinship terminologies from a historical perspective. The underlying ideas and the basic notions of componential analysis are introduced, as described by the pioneers of the field, and some alternative approaches to describing the semantics of kinship terms are briefly sketched.

1.1 General

Every known human language has a kinship terminological system, but different languages have different organizations of these terminological systems; hence the interest of linguists and anthropologists in studying these systems. In his pioneering book Systems of Consanguinity and Affinity of the Human Family Henry Lewis Morgan (1871) made extensive studies of kinship terminologies of the world languages and their reflection in the social structure of society, and this work was extended and enriched by other scholars, notably anthropologist George Peter Murdock (1949). In this tradition, the meaning of kin terms in foreign languages is represented by a primitive English term (for instance ‘mother, ‘father’, etc.), a relative product of two or more primitive terms (for instance ‘mother’s father’) or a collection of primitive and/or relative product terms, where each primitive term and each relative product denotes a ‘kin type’. This type of notation alone, useful as it is for constructing typologies of kinship terms (such as Hawaiian, Eskimo, Crow, etc. already discovered by Morgan), poses certain difficulties to the analyst regarding the important question of what the common pieces of meaning of all the kin types are that allow them to be covered by a single kin term, or what the principles of classification are of kinship in the society the anthropologist/linguist is studying. This question is addressed by ‘componential analysis’, a formal procedure developed in linguistics for other purposes, and based on the Saussurian idea of ‘linguistic system’.

In his Course in General Linguistics (Cours de linguistique générale, 1916), Ferdinand de Saussure created a general linguistic theory at the heart of which lay the notion of ‘linguistic system’ (see the brief review in Pericliev 2010: 2–3). Language (langue), according to Saussure, is a system in the sense that the meaning or value (valeur) of all linguistic entities can only be determined by their contrasts, or distinctions, from all other entities in the same system. ‘In the language itself, there are only differences’, wrote Saussure (1996[1916]: 118; italics in original), ‘A linguistic system is a series of phonetic differences matched with a series of conceptual differences’ (p. 118). The basic task of linguistics, then, is to reveal the structure of linguistic systems by applying the structural method of contrasts and oppositions.

Saussure’s idea of language as a system broke a long tradition in Western thought dominant from Plato on, of viewing language as just an inventory of names (whatever they stood for, ideas or things in the external world), and the goal of language science as relating these names (whether derived from the true nature of things or by convention) to ready-made ideas and things given in advance of language. Saussure, in contrast to this view, conceived language not as a mere collection of discrete items, but as a highly organized totality (or, a Gestalt), in which the items are interrelated and derive their meaning from the system as a whole. Thus, he writes:

In all these cases, what we find, instead of ideas given in advance, are values emanating from the linguistic system. If we say that these values correspond to certain concepts, it must be understood that the concepts in question are purely differential. That is to say they are concepts defined not positively, in terms of their content, but negatively by contrast with other items in the same system. What characterises each most exactly is being whatever the others are not. [Saussure 1996[1916]: 115; italics in original]

The importance of the idea of language as a system, and structuralism in general, cannot be overstated. It influenced researchers both in linguistics and outside of linguistics. Within linguistics, the structural method came to be recognized as an indispensable tool at all levels of linguistic analysis: phonology (‘distinctive feature analysis’), semantics (‘structural semantics’), morphology, etc. Transformationalists (especially Chomsky) emphasized that formal, generative grammar as a whole is a ‘systemic notion’ in that a simplification in some component leads to more complexities in another component. Also, the idea gave rise to different linguistic trends like the Prague school (Trubetzkoy, Jakobson), the Copenhagen school (Hjelmslev) and American structuralism (Bloomfield, Bloch, Harris, etc.). Outside of linguistics, the principles and methods of structuralism were adopted by scholars of such diverse areas as anthropology (Claude Lévi-Strauss), psychoanalysis (Lacan) and literary criticism (Barthes) and were implemented in their respective areas of study. According to Assiter (1984), there are four common ideas regarding structuralism that form an ‘intellectual trend’. First, the structure is what determines the position of each element of a whole. Second, structuralists believe that every system has a structure. Third, structuralists are interested in ‘structural’ laws that deal with coexistence rather than changes. And, finally, structures are the ‘real things’ that lie beneath the surface or the appearance of meaning.

The Saussurian idea of system, and the related formal procedures developed in linguistics for discovering oppositions in phonological and semantic systems, were transferred by direct analogy to componential analysis of kinship terminologies. The pioneers in the field, Floyd Lounsbury and Ward Goodenough, readily acknowledge this. Thus, in a seminal article in Language, Lounsbury states that ‘The aim of this paper is to point out a relatively simple problem in semantics which can be analysed by means of techniques analogous to those already developed in linguistics [ . . . ]’ (Lounsbury 1956: 158–9). Goodenough, analogously, refers to the utility and rigour of the procedures already developed in linguistics:

Inspiration [ . . . ] has come largely from accomplishments of linguistic science. Linguists are able to produce elegant and accurate representations of what one has to know in phonology and grammar if one is to speak particular languages acceptably by native standards. Their procedures enable them to replicate one another’s work readily. Application of the basic strategies of descriptive linguistics to the problem of describing other facets of culture is helping to raise the standards of rigor in ethnographic description. These strategies include what is best described as contrastive analysis. Its use for describing how people classify phenomena, insofar as their classifications are reflected in the vocabulary of their language, has led to the analytic method described here. (Goodenough 1967: 1203)

The following section describes the method of componential analysis of kinship vocabularies in more detail.

1.2 Componential analysis of kinship terminological systems: the basic notions

The method of componential analysis was introduced into kinship semantics basically through the work of anthropologists Lounsbury and Goodenough (Goodenough 1956, 1964, 1967; Lounsbury 1956, 1964, 1965; but see also Greenberg 1949; Wallace and Atkins 1960; Hammel 1965; Leech 1974, and more recently, Geeraerts 2010; Bernard 2011).

1.2.1 Kin terms and kin types

The kin terms of a language, such as English mother, aunt, son-in-law, etc., are linguistic labels for a range of kin types (= denotata), which specify the genealogical position of one’s kin with respect to oneself. In the following, we shall use the following standard abbreviations (Murdock 1949) of atomic genealogical relationships in terms of which the kin types are expressed:

Fa = ‘father’, Mo = ‘mother’, Br = ‘brother’, Si = ‘sister’,

So = ‘son’, Da = ‘daughter’, Hu = ‘husband’, and Wi = ‘wife’.

(Another common notation for the atomic relationships is: F = father, M = mother, B = brother, Z = sister, S = son, D = daughter, H = husband, W = wife.)

Additional symbols may be used to specify relative age or sex of the speaker, for instance:

y = younger, e = elder; m = male ego, f = female ego

These atomic relationships are juxtaposed to express more distant kin types (relatives), as for example, MoBr ‘mother’s brother’, MoSi ‘mother’s sister’, MoSiHu ‘mother’s sister’s husband’, MoBre ‘mother’s elder brother’, etc.

The meaning of kin terms is represented by all kin types, or relatives, covered by the term. For example, the (simplified) meaning of the English term uncle is FaBr and MoBr. The set of all kin terms in a language is the kinship vocabulary of the language.

In the terminology of Charles Morris, used by Lounsbury and Goodenough, the kin types are known as the denotata of a kin term, and the range of meaning, or the class of all denotata of a kin term, are known as the designatum of the term. The (componential) definition of a kin term, in this terminology, is often referred to as the significatum of the term, and I turn to this notion in the next subsection.

1.2.2 Componential definitions of kin terms

A kin term may be viewed as a class and the kin types, covered by the term, as elements of this class. In principle, there are two ways to define a class: either by enumerating the elements of the class, or by giving the ‘defining features’ of the class, that is to say, the necessary and sufficient conditions for membership in the class. This is analogous to the procedures in phonology, where a given phoneme is defined either by listing the allophonic types which belong to it (in which case we are defining it by enumeration), or by a specific bundle of distinctive phonetic features (in which case we are defining it in the second way). The first mode of definition reveals one aspect of the structure of the system, viz. how the kinship semantic field is segmented by the kin types. This definition, however, does not exhibit the underlying principles of organization. This underlying structure can only be revealed if we can proceed by definitions in terms of distinctive semantic features and each kin term is assigned a bundle of features, demarcating each kin term from all remaining kin terms in the kinship vocabulary of a language. Componential analysis is concerned with definitions in terms of distinctive semantic features (or with the significatum of the terms). These are the necessary and sufficient conditions for discrimination.

There is general consensus on this point – for example, Lounsbury writes ‘[the] bundle of features states the necessary and sufficient conditions which an object must satisfy if it is to be a denotatum of the term so defined’ (1964: 1074); see also ‘A significatum is a statement of various necessary and sufficient conditions for a kin type to belong to the class of kin types denoted by a term’ (Wallace and Atkins 1960: 67).

Another essential feature of componential definitions is their con-junctiveness. Such a definition is a Boolean class product, which is a conjunctive (or ‘unitary’) definition. It is assumed in the field that kin terms are susceptible to such representations. Componential analysis is performed starting with definitions of kin terms by their attendant kin types (extensional definitions) and aims to arrive at definitions by distinctive semantic features (intensional definitions). We note that disjunctive definitions would generally be conceived as a major shortcoming by kinship analysts, Lounsbury for example writing on this topic:

We feel that we have failed if we cannot achieve conjunctive definitions for every terminological class in the system. Were we to compromise on this point and admit disjunctive definitions (class sums, alternative criteria for membership) as on a par with conjunctive definitions (class products, uniform criteria for membership), there would be no motivation for analysis in the first place, for definitions of kin classes by summing of discrete members . [ . . . ] . are disjunctive definitions par excellence. (Lounsbury 1964: 1074)

There is also empirical evidence that kin terms are actually conjunctive notions, coming from typological studies. Thus, in a study of grandparent terminology (Greenberg 1966), involving the four kin types FaFa, MoFa, FaMo, and MoMo, fifteen grandparental kin terms are possible, resulting from the different groupings of the kin types. Of these fifteen possible kin terms, it was found, inspecting about one hundred languages, that four did not actually occur in the sample, and these were exactly the kin terms that involved logically disjunctive definitions. Impressive results in this vein were found for sibling terminology as well (Nerlove and Romney 1967; Kronenfeld 1974; Epling, Kirk and Boyd 1973).

1.2.3 Dimensions, or features of contrast

What are the dimensions, or features of contrast, serving to demarcate the kin terms in a kinship vocabulary? There is great diversity in structuring human societies and kinship terminology is, accordingly, diverse. In a widely known paper ‘Classificatory systems of relationship’, Alfred Kroeber (1909) writes: ‘It is apparent that what we should try to deal with is not the hundreds or thousands of slightly varying relationships that are expressed or can be expressed by the various languages of man, but the principles or categories of relationship which underlie these. Eight such categories are discernible.’ He then describes and explains a set of contrasts, or oppositions, existing between kin terms of the languages of the world: (1) differences of generation; (2) the differen...

Table of contents

• Cover
• Title Page
• Copyright
• Contents
• List of tables in the Appendix
• Preface
• Acknowledgments
• 1. Introduction: the historical background
• 2. Problems of componential analysis
• 3. The Kinship system
• 4. Componential analyses of selected languages
• 5.Conclusion
• Appendix: Tables
• Bibliography
• Index